Blockwise Document Metadata Extraction via Ontological Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional document processing technologies face challenges in extracting computational data from scanned ink-on-paper documents due to varied custom formats, individual styles, and diverse alignments, limiting the accessibility and usability of the information they contain.
Innovation Solution
A cognitive document digitization engine that identifies macroblocks and microblocks within document images, employing ontological analysis and confidence levels to extract key-value pairs and output metadata, which includes relative styling parameters, thereby overcoming alignment and semantic relationship dependencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional document processing extracts data from scanned images, then data accessibility is improved, but extraction accuracy deteriorates due to varied custom formats, individual styles, and diverse alignments
Solution Approach 1:
The patent segments the document image processing into hierarchical blocks (macroblocks containing microblocks) to systematically handle varied formats and styles. Each block is processed independently to extract key-value pairs, allowing the system to accommodate diverse alignments and custom formats while maintaining extraction accuracy through structured analysis.
Solution Approach 2:
The patent employs ontological analysis that transforms unstructured visual image data into structured metadata with defined parameters including confidence levels. This parameter transformation enables the system to handle varied document formats by converting diverse visual patterns into standardized data structures with measurable accuracy metrics.
2Measurement precision
If blockwise processing with ontological analysis is implemented, then extraction accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent divides the complex processing task into manageable hierarchical blocks (macroblocks and microblocks), where each block undergoes independent ontological analysis. This segmentation reduces processing complexity by breaking down the overall task into smaller, more manageable units while maintaining high extraction accuracy through systematic analysis of each segment.
Solution Approach 2:
The patent applies local quality analysis by performing ontological analysis specifically on microblock content to extract key-value pairs with confidence levels. This localized approach focuses computational resources on critical extraction points within each block, improving accuracy without requiring complex processing across the entire document uniformly.
3Reliability
If confidence levels are associated with extracted key-value pairs, then data reliability is improved, but processing time increases
Solution Approach 1:
The patent implements partial action by associating confidence levels selectively with extracted key-value pairs based on their extraction certainty. Rather than uniformly processing all data points with equal computational effort, the system focuses ontological analysis on critical blocks and microblocks, improving data reliability for high-confidence extractions while reducing processing time through targeted analysis.
Data Source
AI summary
Methods, computer program products, and systems are presented. The methods include, for instance: obtaining a document image, wherein the document image includes a plurality of objects; identifying a plurality of macroblocks within the document image; performing microblock processing within macroblocks of the plurality of macroblocks, wherein the microblock processing includes examining content of microblocks within a macroblock for extraction of key-value pairs, the examining content including performing an ontological analysis of microblocks, wherein the microblock processing includes associating confidence levels to the extracted key-value pairs; and outputting metadata based on the performing microblock processing within macroblocks of the plurality of macroblocks.